For Tuesday Read Russell and Norvig, chapter 4, section 1 Read - - PowerPoint PPT Presentation
▶
SMART_READER_LITE
LIVE PREVIEW
For Tuesday Read Russell and Norvig, chapter 4, section 1 Read - - PowerPoint PPT Presentation
May 29, 2023 25 likes •236 views
For Tuesday Read Russell and Norvig, chapter 4, section 1 Read Russell and Norvig, chapter 5 Do chapter 3, ex 6 (a, b, and d). Program 1 Late Passes You have 2 for the semester. Only good for programs. Allow you to hand in
SLIDE 1
For Tuesday
Read Russell and Norvig, chapter 4, section
1
Read Russell and Norvig, chapter 5
Do chapter 3, ex 6 (a, b, and d).
SLIDE 2
Program 1
SLIDE 3
Late Passes
You have 2 for the semester.
Only good for programs.
Allow you to hand in up to 5 days late IF
you have a late pass left.
Each good for +.05 on final grade if unused.
Must indicate that you are using a late pass
in Blackboard when you submit.
Only way to turn in late work in this course.
SLIDE 4
Homework
Soccer
Titan
Shopping for AI books
Playing tennis
Practicing tennis
High jump
Knitting
Bidding on an item at an auction
SLIDE 5
Characteristics
Observability
Agents
Deterministic or stochastic
Episodic or sequential
Static or dynamic
Discrete or continuous
Known or unknown
SLIDE 6
Breadth-First Search
List ordering is a queue
All nodes at a particular depth are expanded
before any below them
How does BFS perform?
– Completeness – Optimality
SLIDE 7
Complexity of BFS
Branching Factor
For branching factor b and solution at depth
d in the tree (i.e. the path-length of the solution is d)
– Time required is: 1 + b + b2 + b3 + … bd – Space required is at least bd
May be highly impractical
Note that ALL of the uninformed search
strategies require exponential time
SLIDE 8
Uniform Cost Search
Similar to breadth first, but takes path cost
into account
SLIDE 9
Depth First Search
How does depth first search operate?
How would we implement it?
Performance:
– Completeness – Optimality – Space Complexity – Time Complexity
SLIDE 10
Comparing DFS and BFS
When might we prefer DFS?
When might we prefer BFS?
SLIDE 11
Improving on DFS
Depth-limited Search
Iterative Deepening
– Wasted work??? – What kinds of problems lend themselves to iterative deepening?
SLIDE 12
Repeated States
Problem?
How can we avoid them?
– Do not follow loop to parent state (or me) – Do not create path with cycles (check all the way to root) – Do not generate any state that has already been
generated. -- How feasible is this??
SLIDE 13
Informed Search
So far we’ve looked at search methods that
require no knowledge of the problem
However, these can be very inefficient
Now we’re going to look at searching
methods that take advantage of the knowledge we have a problem to reach a solution more efficiently
SLIDE 14
Best First Search
At each step, expand the most promising
node
Requires some estimate of what is the “most
promising node”
We need some kind of evaluation function
Order the nodes based on the evaluation
function
SLIDE 15
Greedy Search
A heuristic function, h(n), provides an
estimate of the distance of the current state to the closest goal state.
The function must be 0 for all goal states
Example:
– Straight line distance to goal location from current location for route finding problem
SLIDE 16
Beam Search
Variation on greedy search
Limit the queue to the best n nodes (n is the
beam width)
Expand all of those nodes
Select the best n of the remaining nodes
And so on
May not produce a solution
SLIDE 17
Focus on Total Path Cost
Uniform cost search uses g(n) --the path
cost so far
Greedy search uses h(n) --the estimated
path cost to the goal
What we’d like to use instead is
f(n) = g(n) + h(n) to estimate the total path cost
SLIDE 18
Admissible Heuristic
An admissible heuristic is one that never
verestimates the cost to reach the goal.
It is always less than or equal to the actual
cost.
If we have such a heuristic, we can prove
that best first search using f(n) is both complete and optimal.
A* Search
SLIDE 19
Heuristics Don’t Solve It All
NP-complete problems still have a worst-
case exponential time complexity
Good heuristic function can:
– Find a solution for an average problem efficiently – Find a reasonably good (but not optimal) solution efficiently
